agent-evaluation

agent-evaluation is a skill for Claude Code, Codex from sickn33/agentic-awesome-skills. It costs 34 tokens per session (1,551 once invoked), scanned A, original, MIT.

A testing and measurement guide for AI agents. It covers checking what agents can do, how they behave, how reliably they work, and whether changes cause regressions.

In plain words
What is it for?
Use it to design tests and benchmarks, assess capabilities, measure reliability, run regression tests, and monitor agents in production. It does not cover fairness, bias, user experience, or model-training evaluation.
Why use it?
It helps reveal failures that ordinary software tests may miss in systems whose responses can vary. It separates agent behavior testing from model-training measures such as loss and perplexity.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Part of the agentic-awesome-skills plugin — 196 skills shipped together

Good fit Use it to design tests and benchmarks, assess capabilities, measure reliability, run regression tests, and monitor agents in production. It does not cover fairness, bias, user experience, or model-training evaluation.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/sickn33/agentic-awesome-skills/agent-evaluation
About the project

AAS Core is a local control plane for coding agents that lets them search a large catalogue of skills, choose a stack, validate it, and create a reproducible plan. It is used to assemble and review agent workflows through its CLI, local MCP server, catalogue, plugins, and Workbench. The catalogue add-ons provide the skills, plugins, bundles, and workflows that AAS Core helps agents select and validate.

sickn33/agentic-awesome-skills · 46,184 stars · on GitHub · sickn33.github.io

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add sickn33/agentic-awesome-skills --skill agent-evaluation
Clone the repo
git clone --depth 1 https://github.com/sickn33/agentic-awesome-skills

Made for: Claude Code, Codex.

Or install agentic-awesome-skills, the plugin that ships this one along with the rest of its 196 skills.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for agent-evaluation

README.md
[![agentmods](https://agentmods.dev/badge/skills/sickn33/agentic-awesome-skills/agent-evaluation/github.svg)](https://agentmods.dev/skills/sickn33/agentic-awesome-skills/agent-evaluation)
Your own site
<a href="https://agentmods.dev/skills/sickn33/agentic-awesome-skills/agent-evaluation"><img src="https://agentmods.dev/badge/skills/sickn33/agentic-awesome-skills/agent-evaluation/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for agent-evaluation

Your own site · 80×15
<a href="https://agentmods.dev/skills/sickn33/agentic-awesome-skills/agent-evaluation"><img src="https://agentmods.dev/badge/skills/sickn33/agentic-awesome-skills/agent-evaluation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 34 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,551 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • Socket pass 6 Sept 2026
  • Snyk pass 6 Sept 2026
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00034 $0.01551
Opus 5 $0.00017 $0.00776
Sonnet 5 $0.00007 $0.00310
Haiku 4.5 $0.00003 $0.00155

Measured 2d ago against content hash 35a81463e739, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

agent-evaluation scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

Origin

Copies of this mod

8 near-identical copies found in the catalogue:

plugins/agentic-awesome-skills-claude/skills/agent-evaluation/SKILL.md · 87 lines

How it starts

The opening of the file, as written. The whole thing — 87 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Evaluation

Evaluate observable agent behavior against task-specific cases. Modified by AAS maintainers on 2026-09-05 to remove unsupported benchmark claims, correct uncertainty/error reporting and separate optional architecture sketches from the operating procedure.

When to Use

Use when comparing a changed agent, prompt or tool configuration, reproducing an observed failure, or estimating reliability on a declared task distribution. Do not infer product readiness from a public benchmark percentage or a generic score threshold.

Prerequisites

  • A versioned case set with expected observable outcomes and permission boundaries.
  • A known baseline and candidate revision, including model, prompt, tools, configuration and runtime versions.
  • Authorized synthetic or redacted inputs, isolated targets and a bounded token, time and cost budget.
  • A verifier that distinguishes wrong outcomes, expected safe rejections, evaluator failures and infrastructure outages. Provider access is needed only if the declared evaluation calls that provider.

Evaluation procedure

  1. Freeze the contract. Record case IDs and dataset revision, baseline/candidate identities, target environment, repeat plan, budgets, stopping rule and decision criteria before execution. Keep critical safety and authorization failures separate from average quality; they cannot be compensated by a higher score.
  2. Validate the harness. Run a known-pass case, a known-fail case and a deliberate verifier/infrastructure failure. Confirm that each is classified correctly and that trace retention excludes credentials and private input bodies. If classification is wrong, fix the harness and repeat these checks before measuring the agent.
  3. Run the frozen cases. Use the same case definitions and budgets for baseline and candidate, with independent fixture state and recorded execution order. Retain every attempt and its run ID, outcome, reason, latency and resource totals. An exception is not evidence that an unsafe request was safely rejected.
  4. Investigate variation. Preserve the original failure. Classify disagreement as agent behavior, shared-state contamination, verifier ambiguity or an outage. Use only the predeclared repeat budget; do not retry until green, silently drop failures or change the expected outcome to fit the candidate. An unresolved harness fault makes the affected result inconclusive.
  5. Compare and decide. Report per-case results and uncertainty, regressions, critical failures and incomplete cases. Repeated runs of one case are not independent samples of the task distribution. A changed expectation needs a separately reviewed contract revision and reruns of both baseline and candidate; keep the old results.
  6. Fix and verify. Make a bounded fix, rerun the failing case to verify the mechanism, then rerun the applicable frozen regression suite from clean state. Stop at the declared budget if disagreement persists. Record pass, fail or inconclusive with the exact evidence; follow the project publication/deployment approval boundary separately.

Read the full file on GitHub · 87 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago Changed · -1,049 lines · -4 tokens per session scan B → A 35a81463e739
  2. 10d ago First seen · 1,136 lines · 38 tokens per session scan B c7a2bca261ed

Subscribe to this mod's changes

agent-evaluation is a skill published in the GitHub repository sickn33/agentic-awesome-skills (46,184 stars, last pushed yesterday), licensed MIT. It adds 34 tokens to every session and 1,551 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.